Feb. 20, 1963 COMMUNICATIONS TO THE EDITOR 487

useful material for the recovery of a large number of heavy metals which form insoluble sulfides. OAK RIDGE NATIONAL LABORATORY HAROLD 0. PHILLIPS OAK RIDGE, TENN. KURT A. KRAUS

RECEIVED JANUARY 16, 1963

A RAPID AND QUANTITATIVE EXCHANGE OF THE BORON HYDROGENS IN TRIMETHYLAMINE BORANE

WITH Dz0’ Sir :

Close examination of the slow acidic hydrolysis of trimethylamine b ~ r a n e ~ - ~ and the most noteworthy ab- sence of a boron-hydrogen kinetic isotope effect, has demonstrated now that the boron hydride hydrogens are exchanging with the protons of the solvent.

large excess D10 (CH3hNBHa --+ (CHs)aNBDa

DC10.5-5.0 M I fast I1

The homogeneous reaction of I in dilute deuteriochloric acid first produces I1 which can be extracted with ether or allowed to hydrolyze slowly to produce pure deu- terium gas.

Dz0 DCI (CHj)JXBDa __f 3 moles Dz

slow

In keeping with these data the hydrolysis of trimethyl- amine borane-di in hydrochloric acid produces only hydrogen gas.

No other simple derivatives of diborane have been reported to exchange with heavy water.h As I has been used to reduce carbonyl compounds and hydroborate olefins,6-8 the exchange reaction now allows convenient reductive deuteriations using heavy water as the source of deuterium. Norcamphor, benzophenone, cyclo- hexanone and acetone have been reduced to the CY- deuterio alcohols using I1 and boron trifluoride ether- ate.’ Deuterioborations have been performed in re- fluxing toluene.8 Trimethylamine amine borane-d, also has been converted into sodium borodeuteride us- ing sodium rnethoxide in diglyme.

Large amounts of trimethylamine-da are obtained readily using this procedure : sulfuryl chloride (0.50 ml.) is vigorously stirred for twenty minutes with 20 ml. of heavy water. Trimethylamine borane (Callery Chemical Co., 1.000 g., 13.9 mmole) dissolved in fifty ml. of ether is vigorously stirred with the acidic heavy water a t 25’. The exchange reaction is followed by infrared analysis. The extent of hydrolysis is deter- mined manometrically. After six hours the deuterium content of the amine borane is 98% while less than 6 7 , has hydrolyzed. The ether layer is dried over potas- sium ~ a r b o n a t e . ~ Evaporation in vacuo leaves a white residue which is sublimed. The product weighs 0.997 g. (m.p. 9 4 O , undepressed with I, 98.1% deuterium by mass spectrum). (1) Boron Hydrides. VI. Paper V, R. E. Davis and J. A. Gottbrath,

(2) G. E. Ryschkewitsch, ibid., 82, 3290 (1960). (3) R. E. Davis, C. L. Kibby and C. G. Swain, ibid., 83, 5950 (1960). (4) R. E. Davis, ibid. , 84, 892 (1962). (5) Decaborane will produce BlOHlOD4: G. A. Guter and G. W. Schaeffer,

ibid., 78 , 3340 (1950); K. Atterbrrry, J. Phys . Chcm., 6 2 , 1457 (1958); R. J. F. Palchak, J. H. Norman and R . E. Williams, J . A m . Chem. Soc., 83, 3380 (1961). The BltHiz-2 ion will completely exchange with heavy water (E. L. Muetterties, R. D. hlenitield, H. C. Miller, W. H. Knoth, Jr., and J . R. Downing, ibid., 84, 2506 (1962)). Xeither has found application as a useful reducing agent in organic cheniistry.

(6) E . C. Ashby, ib id . , 81, 4791 (1959). ( 7 ) W. M. Jones, ibid. , 82, 2528 (1960). (8) H . C. Brown, K. J. Murray, L. J. Murray, J. A. Snover and G. Zweifel,

(9) This solution can be used directly since the amount of I1 can be

J . A m . Chcm. Soc., 84, 895 (1962).

ibid., 82, 4233 (1960).

calculated from the amount of gas produced.

These new procedures should greatly facilitate the synthesis of compounds labeled with deuterium or tritium with known stereochemistry to aid in the elucidation of reaction mechanisms.

Acknowledgment.-The present study was sup- ported by a Frederick Gardner Cottrell grant from the Research Corporation and a grant from the Sloan Foundation. Discussion with Prof. R. W. Parry is appreciated.

(10) Alfred P. Sloan Fellow, 1962-1964. (11) This study was begun by Miss A. E. Brown as an undergraduate

(12) Iiational Science Foundation Predoctoral Cooperative Fellow, research problem.

1961-1963.

DEPARTMENT OF CHEMISTRY ROBERT EARL D A V I S ~ ~

LAFAYETTE, INDIANA RIJDOLF HOPMANN CHARLES L. K I B B Y ~ ~

RECEIVED JANUARY 14, 1963

PURDUE UNIVERSITY ANN ELIZABETH BROW’H“

ISOLATION OF ANTIBODY BY MEANS OF AN IMMUNOLOGICAL SPECIFIC ADSORBENT

Sir : As the immunochemist strives for a better under-

standing of the physicochemical factors involved in antigen-antibody interactions, the need for purified reactants becomes apparent. In his search he has devised a number of non-specific and specific techniques to obtain purified preparations of antibody. The non- specific methods are based on the fractionation of serum to provide a purified preparation of y-globulin, and include such methods as ammonium sulfatelJ or alcohol f ra~t ionat ion,~ electrophoresis4 and column chromatography on DEAE-cellulose.6 Although these techniques provide fairly pure y-globulin preparations the specific irrmune y-globulin desired may be only a small percentage of the total protein in the purified preparation. Hence, one obtains a good yield but low purity with respect to antibody.

The specific methods of antibody purification in- volve the removal of antibody by precipitation with the specific soluble antigen or reaction of antiserum with antigen in some insoluble state, and subsequent dis- sociation of the antibody from the precipitate by one of several method^.^-^ Many of these specific methods depend upon some special property (;.e., insolubility of the antigen in high salt concentration, mercurial salts, etc.) of the particular antigen-antibody system and therefore limit their applicability. The more general method of alkaline dissociation of antigen-antibody precipitates often leads to denaturation of the anti- body, and particularily upon prolonged contact with the antibody. The method of acid dissociationlo of antigen-antibody precipitates a t pH 3.0 has been used by many investigators, and seems to be the most appli- cable. However, the soluble protein antigens must be rendered insoluble by some other procedure such as

(1) R. A. Kekwick, Biochcm. J . , 34, 1248 (1940). (2) E. J. Cohn, T. L. McMeekin, J. L. Onclcy, J, M. Newel1 and W. L.

(3) E. J. Cohn, J. A. Luetscher, J. L. Oncley, S. H. Armstrong and B. D.

(4) A. Tiselius, Biochcm. J . , 31, 1464 (1937). (5) H. A. Sober, F. G. Gutter, M. M. Wyckoff and E. A. Peterson, J . A m .

(6) M. Heidelbcrger and F. E. Kendall, J . E x p t l . Mcd.. 64, 161 (1936). (7) M. Heidelberger and E. A. Kabat, ibid. , 67, 181 (1938). (8) L. A. Steinberger and D . Pressman, J . Immunol. , 66, 65 (1950). (9) D . H. Campbell, R. H. Blnker and A. B. Pardee, J . A m . Chcm. Soc.,

7 0 , 2496 (1948). (10) D . H. Campbell and F. Lanni in D. M. Greenberg, “Amino Acids

and Proteins,” 1st ed., Charles C. Thomas, Springfield, Ill., Chapter S I , 1951.

Hughes, J . A m . Chem. Soc., 62, 3386 (1940).

Davis, ibid. , 62, 3396 (1940).

Chein. SOC., 78, 756 (1956).

48s COMMUNICATIONS TO THE EDITOR Vol. 8.5

selective precipitation, l 1 or modification of the anti- gen.8312

An approach to the isolation of antibody against a soluble protein antigen was suggested by Campbell, Leuscher, and Lerman.I3 They coupled a soluble protein antigen to an insoluble, modified (p-amino- benzyl-) cellulose by means of a diazonium bond. This provided an insoluble protein antigen which could specifically combine with antibody and upon acidifi- cation to pH 3.2 be separated by centrifugation into soluble antibody and insoluble antigen. This method provided a yield of 57y0 of the total antibody present in the original serum as determined by quantitative precipitation with specific antigen. The purity (pre- cipitable antibody/total protein) of the soluble material recovered by specific adsorption and subsequent elution was 90%.

This problem of immune adsorbents was re-investi- gated in our Laboratory as a result of commercially available p-aminobenzylcellulose, Cellex-PAB, (Bio- Rad Lab., Richmond, Calif.) and the coupling of a simple haptenic material, arsanilic acid, to cellulose (ICN Corp., City of Industry, Calif.). The method employed in our study involves the use of an antigen “fixed” by a diazo bond to the commercially available cellulose to which the p-aminobenzyl group has been coupled. After exposure of the appropriate antiserum to the antigen-coupled Cellex, the column was washed prior to elution of antibody with a glycine-HC1 buffer pH 3.0. Utilizing this method we have isolated rabbit antibodies to (1) bovine serum albumin (BSA), ( 2 ) whole ragweed pollen extracts (WRE), (3) a highly purified Timothy pollen fraction1* and (4) a simple hapten (arsanilate) .

In order to couple the protein antigens to the Cellex- PAB diazotization was carried out a t 0’ with 2 N HC1 and 14y0 NaNOz for 1 hour. The diazotized Cellex-PAB was filtered and washed successively with 5% sodium acetate, 5% urea, and distilled water. Coupling of the antigen was accomplished by addition of the diazotized Cellex to the protein solution at pH 8, and the suspension was magnetically stirred overnight a t 2’. The non-reacting antigen was re- covered by filtration, and the Cellex was further coupled with ,?-naphthol to block any unreacted free diazonium sites. The antigen-coupled Cellex was washed SUC- cessively with distilled water, glycine-HC1 buffer, pH 3.0, and was then readjusted to pH 7.25 with 0.1 -V NaOH. The immune adsorbent then was poured as a slurry into a 1 x 10 cm. column, packed by gravity a t 2’, and washed with an appropriate buffer (pH 7.25).

The isolation of purified rabbit anti-whole ragweed pollen extract (AWRE) antibodies from a whole rag- weed-coupled Cellex (WRE) column was accomplished in the following manner. To 1 g. samples of \VRE coupled-Cellex was added 20 ml. of rabbit immune serum containing 50 mg. of antibody protein precipi- table by WRE. The time required for the serum to pass through the column was 1-2 hours. The column was then washed with 50-100 ml. of a citrate-phosphate buffer, pH 7.25, in order to remove all non-specific proteins. The specifically bound antibody then was dissociated from the immune adsorbent with a glycine- HCl buffer, pH 3.0. Five 5-ml. fractions were col- lected, and adjusted to pH 7.4 with 0.1 N NaOH. The presence of antibody in the collected fractions was

(11) S. I. Epstein, P. Doty and ‘A’. C . Boyd, .I. A m . Chem. Soc., 78, 330G ( 1 9 3 )

(12) S. J. Singer, J. E. Fothergill and J. R . Shainoff, ibid., 81, 2277 (1959).

(13) D. H. Campbell, E. Leuscber and I,. S. Lerman, Proc. Nnl l . A r d .sci. 1:. s., 37, 575 (1951).

(14) A. Malley, A. Lietze and C. E. RePd, J. Allergy, 33, 84 (19A2).

determined by ring tests and a quantitative precipitin test. The recovery of protein from the eluted frac- tions varied from 24-35 nig., which represented a yield of 46-707,. and 83Yc of the recovered protein was precipitable with specific antigen.

The same general procedure was used in the isolation of rabbit antibodies directed against bovine serum albumin, arsanilate hapten, and a highly purified Timothy pollen fraction. I t also has been applied to the isolation oi human reaginic antibody from non- treated Timothy sensitive patients, l6 and these results will be presented in more detail in a subsequent publi cation.

This work was supported by a grant from the U. S. Public Health Service; Grant No. 41-01339.

COSTRIBUTION h’o. 2920 ARTHUR MALLEY DIVISION OF CHEMISTRY AXD CHEMICAL ENCINEERIXG CALIFORNIA INSTITUTE OF TECHNOLOGY PASADENA, CALIFORNIA DAS H. CAMPBELL

RECEIVED DECEMBER 19, 1962

(15) A. Malley, Federation PYOC., 21, No. 2 , 14 (1962).

NON-STEROID ANTI-INFLAMMATORY AGENTS Sir:

We wish to report a new class of anti-inflammatory and antipyretic agents, substituted indole acetic and propionic acids. Of some three hundred and fifty indole derivatives studied, one member of the series, 1 - (p chlorobenzoyl) -5-methoxy-2-methylindole-3-acetic acid (I), designated as indomethacin, has demonstrated a high degree of anti-inflammatory activity both in the granuloma inhibition assay’ and in the foot-edema

c=o I Q

CI I

assay.? Indomethacin was equally active in intact or adrenalectomized animals and was also active by local application directly to the cotton pellets. Its anti-inflammatory potency relative to phenylbutazone varied from ten to eighty-five times depending on the test employed.3 Fever induced in rabbits by intra- venous injection of pyrogenic lipopolysaccharide pre- pared from E . coli was effectively blocked by I sub- cutaneously administered. Antipyretic potency of I was approximately 10 times that of aminopyrine and 20 times phenylbutazone, with duration of action resembling the latter compound. In animals indo- methacin is relatively free of activities referable to central nervous, autonomic or cardiovascular systems.

For the synthesis of indomethacin, 5-niethoxy-2- methylindole-3-acetic acid4 was converted to its anhy- dride with dicyclohexylcarbodiimide in tetrahydrofuran. The anhydride was treated with zinc chloride and t - butanol to give t-butyl 5-methoxy-2-methyl-3-indolyl- acetate, m.p. 110-11l0. Acylation of the t-butyl ester with p-chlorobenzoyl chloride afforded t-butyl 1-(p-

(1) R . Meier, W. Schuler and P. Desaulles, E x p e r i e d a , 6, 469 (1950). as modified by C. A. Winter and C. C. Porter, J. A m . Pharni. Assoc., Sei. Ed. , 46, 515 (1957).

(2) C. A Winter, E. A. Risley and G. W, S u s s , forthcoming publication<. (3) Preliminary clinical reports of indomethacin indicate tha t its potmcy

( 4 ) E. Shaw, J . Am, Chem. Soc., 77, 4319 (1955). in nian is about 0.25-0.R tha t of cortisone.